Synthetic Surfactant Protein B Peptides Antimicrobial Activity Of

Home , Melittin, Surfactant protein B

Antimicrobial Activity of Native and Synthetic Surfactant Protein B Peptides Marnie A. Ryan, Henry T. Akinbi, Alicia G. Serrano, Jesus Perez-Gil, Huixing Wu, Francis X. McCormack and This information is current as Timothy E. Weaver of September 27, 2021. J Immunol 2006; 176:416-425; ; doi: 10.4049/jimmunol.176.1.416 http://www.jimmunol.org/content/176/1/416 Downloaded from

References This article cites 30 articles, 7 of which you can access for free at: http://www.jimmunol.org/content/176/1/416.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 27, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Antimicrobial Activity of Native and Synthetic Surfactant Protein B Peptides1

Marnie A. Ryan,* Henry T. Akinbi,* Alicia G. Serrano,‡ Jesus Perez-Gil,‡ Huixing Wu,† Francis X. McCormack,† and Timothy E. Weaver2*

Surfactant protein B (SP-B) is secreted into the airspaces with surfactant phospholipids where it reduces surface tension and prevents alveolar collapse at end expiration. SP-B is a member of the saposin-like family of proteins, several of which have antimicrobial properties. SP-B lyses negatively charged liposomes and was previously reported to inhibit the growth of Escherichia coli in vitro; however, a separate study indicated that elevated levels of SP-B in the airspaces of transgenic mice did not confer resistance to infection. The goal of this study was to assess the antimicrobial properties of native SP-B and synthetic peptides derived from the native peptide. Native SP-B aggregated and killed clinical isolates of Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus, and group B streptococcus by increasing membrane permeability; however, SP-B also lysed Downloaded from RBC, indicating that the membranolytic activity was not selective for bacteria. Both the antimicrobial and hemolytic activities of native SP-B were inhibited by surfactant phospholipids, suggesting that endogenous SP-B may not play a signiﬁcant role in alveolar host defense. Synthetic peptides derived from native SP-B were effective at killing both Gram-positive and Gram-negative bacteria at low peptide concentrations (0.15–5.0 ␮M). The SP-B derivatives selectively lysed bacterial membranes and were more resistant to inhibition by phospholipids; furthermore, helix 1 (residues 7–22) retained signiﬁcant antimicrobial activity in the presence of native surfactant. These results suggest that the role of endogenous SP-B in host defense may be limited; however, http://www.jimmunol.org/ synthetic peptides derived from SP-B may be useful in the treatment of bacterial pneumonias. The Journal of Immunology, 2006, 176: 416–425.

he respiratory tree terminates in small sac-like structures SP-B in both mice and humans results in lethal neonatal respira- (alveoli) that provide an extensive gas exchange surface tory distress syndrome (1, 2). T composed of type I epithelial cells. Hydration of the gas In addition to its biophysical function, surfactant plays an im- exchange surface leads to elevated surface tension at the air/liquid portant role in maintaining the sterility of the gas exchange sur- interface, generating a high collapsing force at end expiration.

face. The surface film serves as a physical barrier to inhaled patho- by guest on September 27, 2021 Type II epithelial cells synthesize and secrete pulmonary surfac- gens and the hydrophilic surfactant proteins SP-A and SP-D, tant, which forms a stable phospholipid-rich film at the air/liquid associated with the large and small aggregate fractions of surfac- interface and prevents alveolar collapse and impaired gas ex- tant, respectively (3, 4), promote clearance of microorganisms 3 change. Dipalmitoylphosphatidylcholine (DPPC), the main lipid from the distal airspaces. SP-A and SP-D opsonize, aggregate, and component of surfactant, reduces surface tension to near zero as enhance phagocytosis of microbes by resident macrophages (5–7). the surfactant film is compressed during exhalation. During inha- SP-A and SP-D may also directly kill bacteria by permeabilizing lation, surfactant phospholipids are inserted into the expanding the bacterial membrane (8). The role of SP-B in alveolar host surface film, a process facilitated by the hydrophobic peptides sur- defense is less clear. A synthetic peptide corresponding to SP-B factant protein B (SP-B) and SP-C. The importance of SP-B for was reported to inhibit the growth of Escherichia coli in vitro (9); surfactant function is underscored by the fact that deficiency of however, bacterial burden was not increased in the lungs of SP-B heterozygous null (SP-Bϩ/Ϫ) mice nor was protection conferred by increased expression of SP-B in transgenic mice (10). *Division of Pulmonary Biology, Cincinnati Children’s Hospital Medical Center, and SP-B is a member of the saposin-like family of proteins University of Cincinnati College of Medicine, and †Division of Pulmonary and Crit- ical Care Medicine, Department of Internal Medicine, University of Cincinnati Col- (SAPLIP), several members of which exhibit potent antimicrobial lege of Medicine, Cincinnati, OH 45229; and ‡Departamento de Bioquimica y Bio- activity (11). SAPLIP family members NK-lysin, granulysin, and logia Molecular I, Facultad Biologia, Universidad Complutense, Madrid, Spain amoebapore all kill bacteria by permeabilizing bacterial mem- Received for publication June 28, 2005. Accepted for publication September 29, 2005. branes, but the mechanism of membrane permeabilization differs among the peptides. Positively charged amino acids located on the The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance surface of NK-lysin and granulysin mediate interaction of the pep- with 18 U.S.C. Section 1734 solely to indicate this fact. tides with the negatively charged membranes of bacteria, resulting 1 This work was supported by National Institutes of Health Grant R37-HL56285 (to in membrane destabilization and/or permeabilization (12, 13). In T.E.W.). J.P.-G was supported by a grant from the Spanish Ministry of Science and Education (BIO2003-09056). F.X.M. was supported by HL68861 and a Veterans contrast, amoebapore A is much more hydrophobic and permeabi- Affairs Merit Award. lizes bacterial membranes in a pH- and oligomerization-dependent 2 Address correspondence and reprint requests to Dr. Timothy E. Weaver, Cincinnati manner (14). SP-B shares features with both types of SAPLIP Children’s Hospital Medical Center, Division of Pulmonary Biology, 3333 Burnet antimicrobial peptides: it is very hydrophobic and forms oligomers Avenue, Cincinnati, OH 45229-3039. E-mail address: [email protected] similar to amoebapore but is also cationic (net positive charge of 3 Abbreviations used in this paper: DPPC, dipalmitoylphosphatidylcholine; PG, phos- ϩ phatidylglycerol; SP-B, surfactant protein B; SAPLIP, saposin-like family of proteins; 7) and lyses negatively charged liposomes at neutral pH, similar h, human; BALF, bronchoalveolar lavage fluid; N-term, N-terminal. to NK-lysin and granulysin (15, 16). We have previously mapped

the lytic domain of SP-B to helix 1, an ␣-helical, amphipathic were incubated with liposomes at 20:1 or 10:1 lipid:peptide ratios for 5 min region containing a net charge of ϩ3 (16). In the present study, before adding to bacteria. Methanol controls were included in all native SP-B and lytic peptides derived from the native peptide experiments. were tested for antimicrobial activity against both Gram-positive Bacterial clearance in transgenic mice and Gram-negative bacteria. Wild-type mice and transgenic mice, in which SP-B concentrations in the airspaces were increased 2- to 3-fold (10) were anesthetized with isoflu- Materials and Methods rane, and a dose of 104 CFU of K. pneumoniae suspended in 100 ␮lof Materials sterile PBS was delivered intratracheally just beneath the cricoid cartilage as previously described (10). To assess bacterial clearance, mice were DPPC and phosphatidylglycerol (PG) were purchased from Avanti Lipids. anesthetized 24 h postinfection with i.p. pentobarbital, exsanguinated by HEPES buffer was purchased from Cambrex Bioscience, and melittin pep- transecting the abdominal aorta, and lung and splenic tissues were har- tide from bee venom was purchased from Sigma-Aldrich. vested and subsequently homogenized in sterile PBS. Serial dilutions of Peptide design homogenates were plated on blood agar plates and incubated overnight at 37°C. Colony counts in the lung and spleen (data not shown) were obtained Synthetic peptides were designed to the proposed helices and interhelical and expressed as CFU/gram of tissue. loops of the mature SP-B peptide (16). Peptides were synthesized by Bio- synthesis Inc. by F-moc chemistry and purified to Ͼ95% homogeneity by Hemolytic assay HPLC. Peptide composition was confirmed by mass spectrometry. Stock solutions (1 mg/ml) were prepared in methanol and diluted into assay Fresh human RBC (hRBC) were rinsed in PBS and centrifuged for 10 min buffer to achieve the peptide concentrations indicated in the figures. Ap- at 800 ϫ g three times and resuspended in PBS to a final erythrocyte ␮ propriate solvent controls were used in each experiment. concentration of 4% v/v. The hRBC suspension (100 l) was added to a Downloaded from 96-well microtiter plate and incubated with individual SP-B peptides (2 Preparation of native human SP-B mg/ml stocks dissolved in methanol) at 2.5, 5.0, and 10.0 ␮M. Controls for zero and 100% hemolysis consisted of hRBC suspended in PBS and 1% Human SP-B was isolated from bronchoalveolar lavage fluid (BALF) of Triton X-100, respectively; additional controls included hRBC suspended patients with pulmonary alveolar proteinosis, as described by Shen et al. in PBS containing 0.5 or 1% methanol. The hRBC/peptide suspension was (17). Briefly, surfactant was isolated from BALF by centrifugation and incubated with agitation for 60 min at 37°C. The samples were centrifuged dissolved in chloroform/methanol (2:1). The organic phase was recovered, at 800 ϫ g for 10 min, and the release of hemoglobin was monitored by dried, dissolved in chloroform/methanol/0.1 M HCl (1:1:0.1 (v/v)) and measuring the absorbance of the supernatant at 550 nm. http://www.jimmunol.org/ loaded onto an LH-60 Sephadex column equilibrated in the same solvent system. Fractions eluted from the column were screened by SDS-PAGE Bacterial aggregation assays and silver staining. SP-B-containing fractions were recovered and dialyzed

(SnakeSkin dialysis tubing; m.w. cutoff, 3500; Pierce Chemical) against Bacteria were grown until mid-log phase, diluted to an OD600 of 0.1, and chloroform/methanol (2:1 (v/v)) overnight at 4°C to remove HCl and was plated in a 96-well polystyrene plate. Native SP-B or synthetic peptides in quantitated by amino acid composition analysis (18). methanol were added to bacteria and incubated at 37°C for 3 h. Samples were stained using the permeant fluorescent probe Syto 9 and impermeant In vitro bacterial killing assay fluorescent probe propidium iodide (BacLight Bacterial Viability kit; Mo- Clinical isolates of Klebsiella pneumoniae (KPA1 serotype), Staphylococ- lecular Probes). Bacteria were examined by fluorescence microscopy to assess bacterial aggregation and changes in propidium iodide or Syto 9 cus aureus,orPseudomonas aeruginosa were grown in Luria broth (LB) by guest on September 27, 2021 and group B streptococcus (provided by J. Wright, Duke University Med- staining compared with untreated or methanol-treated controls. ical Center, Durham, NC) was grown in Todd Hewitt broth at 37°C with continuous shaking to exponential phase. The bacteria were harvested from Isolation of pulmonary surfactant ϫ broth by centrifugation at 500 g for 10 min, washed, and resuspended in Surfactant was isolated by high-speed centrifugation of cell-free BALF (2 3 ␮ sterile PBS at a concentration of 10 CFU/100 l. The concentration of ml in sterile PBS/mouse) obtained from 25-g FVBN mice (6–8 wk old). bacteria was verified by quantitative culture on sheep blood agar plates. Phosphorous in total BAL was measured by the Bartlett assay (19). One hundred microliters of bacterial suspension was plated in a 96-well polystyrene microtiter plate (BD Biosciences), and serial dilutions of each peptide in methanol were added to individual wells in triplicate and incu- Data analysis bated for6hat37°C with rocking. Bacteria were subsequently dispersed, All data are expressed as mean Ϯ SEM. Differences between groups were and aliquots were plated on blood agar plates to obtain colony counts. determined by ANOVA followed by Student-Newman-Keuls or Dunnett Viable pathogen counts after peptide treatment were determined from the posttests if p Ͻ 0.05. Differences between two groups were determined by number of colonies obtained on the methanol-treated control plates com- Student’s t test. pared with the number of colonies from peptide-treated samples. Bacterial killing results are expressed as follows: Percent killing ϭ 100 ϫ (CFU from control wells (without SP-B or peptide) Ϫ CFU from experimental Results wells)/(CFU from control wells (without SP-B or peptide)). Antimicrobial activity of human SP-B against K. pneumoniae Bacterial membrane permeabilization assay Sequence alignments revealed that the location of the cysteine residues in SP-B was a common feature among SAPLIP family mem- K. pneumoniae was grown to mid-log phase in Luria broth at 37°C, washed bers, several of which are bacteriolytic (20). To determine whether twice with 5 mM Tris and 150 mM NaCl, and diluted to OD600 1.0. Bac- teria were aliquoted into a 96-well polystyrene microtiter plate to a final SP-B was also bacteriolytic, a clinical isolate of K. pneumoniae 3 concentration of OD600 0.5. Increasing concentrations of native SP-B (0.1– (10 CFU) was incubated with increasing concentrations of puri- 5.0 ␮g/ml final concentration) were added to each well and incubated for fied human SP-B for6hat37°C. Mature SP-B peptide exhibited 15 min at 37°C with shaking. Alkaline phosphatase enzyme levels were A quantitated over a 90-min period in the presence of the fluorescently la- potent, dose-dependent antimicrobial activity (Fig. 1 ), killing beled enzyme substrate ELF-97 (Molecular Probes) at excitation and emis- Ͼ90% of K. pneumoniae at a concentration of 1.0 ␮M. Incubation sion wavelengths of 355 and 535 nm, respectively. of SP-B with bacteria also resulted in dose-dependent detection of the bacterial periplasmic enzyme alkaline phosphatase, consistent Preparation of phospholipid vesicles with membrane permeabilization (data not shown).

Phospholipids in chloroform were dried under N2 and resuspended in 50 Because SP-B is always associated with membranes, experi- mM HEPES/140 mM NaCl/0.1 mM EDTA buffer (pH 7.0) to a final con- ments were designed to determine whether SP-B-mediated bacte- centration of 1 mg/ml. The phospholipid suspension was passed through a miniextruder (Avanti Lipids) at 45°C through two stacked 0.1-␮m poly- rial killing was altered in the presence of surfactant phospholipids. carbonate filters. A series of 10 extrusions was performed to generate a Surfactant-like liposomes, composed of DPPC/PG (9:1, w/w) or population of unilamellar liposomes with diameters of ϳ100 nm. Peptides DPPC at a 20:1 lipid:protein ratio, were first mixed with SP-B 418 ANTIMICROBIAL ACTIVITY OF SP-B

FIGURE 1. Effect of SP-B on K. pneumoniae viability. A, Increasing amounts of native hSP-B dissolved in methanol were added to K. pneumoniae (103 CFU) in 100 ␮l of sterile PBS and incubated for6hat37°C. Results (n ϭ 4) are expressed as mean Ϯ ,p Ͻ 0.05 vs methanol control. B ,ء ;SEM SP-B was preincubated with liposomes (20:1 lipid:peptide ratio) composed of DPPC or DPPC/PG (9:1) and then incubated with K. pneumoniae. Bacteria were dispersed, plated on blood agar plates, and incubated overnight at 37°C. Solvent (methanol) controls were included in all experiments. Results ,ء ;n ϭ 6) are expressed as mean Ϯ SEM) p Ͻ 0.05 vs SP-B lipid free; ϩ, p Ͻ 0.05 vs SP-B/DPPC/PG. C, Wild-type and transgenic mice expressing elevated levels of hu- Downloaded from man SP-B were intratracheally instilled with 104 CFU of K. pneumoniae, and lungs were harvested 24 h postinfection. CFU were counted from plated lung homogenates, and data are expressed as CFU per gram of lung tissue Ϯ SEM; WT vs transgenic SP-B over- expressors, p ϭ 0.7363. n ϭ 8 mice/group. http://www.jimmunol.org/

followed by incubation with bacteria. Surfactant phospholipids de- To determine whether surfactant phospholipids altered the abil- creased SP-B-mediated killing by ϳ70% (Fig. 1B); however, re- ity of SP-B to aggregate bacteria, SP-B was added to DPPC or moval of PG from the liposomes partially restored activity, result- DPPC/PG liposomes before incubation with bacteria (OD600 0.1) ing in only a 30% decrease in bacterial killing. We previously for 90 min. The presence of DPPC liposomes did not affect the by guest on September 27, 2021 reported that killing of P. aeruginosa and group B streptococcus ability of SP-B to induce aggregation or alter the number or size of was not enhanced in transgenic mice in which the concentration of bacterial aggregates (data not shown). SP-B-induced bacterial ag- SP-B in BALF was increased 2- to 3-fold (10). To determine gregation also occurred in the presence of DPPC/PG liposomes but whether this outcome was related to lipid inhibition or pathogen- to a lesser extent. Propidium iodide staining was reduced in the specific effects of SP-B, K. pneumoniae (104 CFU) were intratra- presence of both DPPC and DPPC/PG lipids consistent with de- cheally instilled into transgenic mice, and bacterial burden was creased CFU in the bacterial killing assays. Thus, surfactant phos- assessed 24 h postinfection (Fig. 1C). Killing of K. pneumoniae pholipids, in particular PG, inhibited both bacterial killing and was not enhanced in transgenic mice, supporting the hypothesis aggregation. that surfactant phospholipids inhibit the bactericidal activity of native SP-B in vivo. Hemolytic activity of human SP-B To determine the specificity of the membranolytic activity of Effect of SP-B on bacterial aggregation SP-B, native peptide was incubated with hRBC in the presence or The hydrophilic surfactant proteins SP-A and SP-D play important absence of DPPC or DPPC/PG liposomes (Fig. 3). SP-B induced roles in lung host defense by inducing bacterial aggregation. To a dose-dependent release of hemoglobin from RBC at a concen- determine whether SP-B could also induce bacterial aggregation, tration of 1.0–7.5 ␮M. Membrane lysis was significantly reduced

K. pneumoniae (OD600 0.1) was incubated with SP-B for 90 min in the presence of DPPC and was virtually ablated in the presence and stained with the vital dyes Syto 9 (green fluorescence indicates of DPPC/PG liposomes. live bacteria) and propidium iodide (red fluorescence indicates dead/dying bacteria). Bacteria were examined by fluorescence mi- Antimicrobial activity of SP-B synthetic peptides against K. croscopy to assess bacterial aggregation and to detect changes in pneumoniae propidium iodide or Syto 9 staining compared with untreated or To map the antimicrobial domain(s) of SP-B, synthetic peptides methanol-treated controls. Addition of SP-B (1–3 ␮M) to K. pneu- were made to the proposed helices and interhelical loops of human moniae induced significant bacterial aggregation compared with SP-B based on the three-dimensional structure of NK-lysin, as controls (Fig. 2). The mean area of bacterial aggregates was 540 Ϯ previously described (16) (Fig. 4A). Antimicrobial activity was 80 ␮m2, and aggregates as large as 5000 ␮m2 were detected. Sim- assessed by incubating individual synthetic peptides with a clinical ilar results were obtained with other strains of bacteria including P. isolate of K. pneumoniae (103 CFU) for 6 h at 37°C. Bacteria were aeruginosa, S. aureus, and group B streptococcus (data not subsequently plated on blood agar plates, and the number of col- shown). Increased propidium iodide staining was detected in SP- onies was counted after 18 h. SP-B peptides containing helix 1 B-treated samples but not in untreated or vehicle controls, indicat- exhibited potent antimicrobial activity against K. pneumoniae (Fig. ing that the aggregated bacteria were also killed. 4B). A peptide encompassing residues 1–37 (N-terminal (N-term) The Journal of Immunology 419

FIGURE 2. SP-B-mediated bacterial aggregation and membrane permeabilization. K. pneumoniae

(OD600 0.1) were incubated with human SP-B (1–3 ␮M) or solvent (methanol) controls for 90 min at 37°C. Bacteria were stained with the vital dyes Syto 9 (stains live bacteria) (A, C, D, G, and J) and propidium iodide (stains dead/dying bacteria) (B, E, F, H, and K) and analyzed by ﬂuorescence microscopy to assess aggre- Downloaded from gation and viability. Results are representative of four separate experiments and depict small (G–I) and large aggregates (J–L). Bar, 20 ␮m. http://www.jimmunol.org/ by guest on September 27, 2021

helix 1,2) killed Ͼ60% of the bacteria at a concentration of 2.5 Dose response of SP-B peptides on bacterial killing ␮M. Removal of the N-terminal 6 aa from N-term helix 1,2 resulted in significantly higher levels of bacterial killing (Ͼ80%). To determine the lowest concentration of SP-B peptide required Helix 1 (residues 7–22) killed Ͼ80% of the bacteria as did a for K. pneumoniae killing, dose-response curves were generated shorter helix 1 peptide (residues 10–22) (data not shown). In con- for the most effective synthetic peptides (helix 1 (residues 7–22), trast, helix 2 and a peptide encompassing helices 3,4,5 exhibited N-term helix 1, and helix 1,2). Helix 1,2 was significantly more much lower levels of antimicrobial activity. These results demon- effective at bacterial killing than helix 1 or N-term helix 1 and strate that residues 10–22 (helix 1) are sufficient for bacterial exhibited significant antimicrobial activity (30%) at concentrations killing. as low as 0.075 ␮M (Fig. 5A). Maximal killing was attained at a

FIGURE 3. Hemolytic activity of native hSP-B. hRBC (4%) were incubated for 1 h with SP-B or SP-B preincubated with DPPC or DPPC/PG (9:1) liposomes (20:1 ratio of lipid: peptide) at 37°C. Methanol-treated controls were included in all experiments and had no effect on hemolysis. Results are the mean of four separate p Ͻ 0.001 vs ,ء ;experiments Ϯ SEM native SP-B lipid-free; ϩ, p Ͻ 0.05 vs SP-B DPPC. 420 ANTIMICROBIAL ACTIVITY OF SP-B

FIGURE 4. SP-B synthetic peptides containing helix 1 kill K. pneumoniae. A, Bars represent synthetic peptides designed to the proposed helical regions (gray) and interhelical loops (black) of the human SP-B mature peptide. The numbers in parentheses represent the corresponding amino acids in the native human SP-B peptide. B, SP-B peptides (2.5 ␮M) or solvent controls were added to K. pneumoniae (103 CFU) in 100 ␮lof sterile PBS and incubated for6hat37°C. Bacteria were dispersed, plated on blood agar plates, and incubated overnight at 37°C to obtain colony counts. Results (n ϭ -p Ͻ 0.05 vs meth ,ء ;are expressed as mean Ϯ SEM (8 anol control. Downloaded from

concentration of 2.5 ␮M for helix 1,2 and 5.0 ␮M for helix 1. To Effect of surfactant phospholipids on SP-B-mediated bacterial killing further characterize the bacteriolytic activity of the SP-B peptides,

increasing concentrations of helix 1 were incubated with K. pneu- Experiments were designed to determine whether SP-B-mediated http://www.jimmunol.org/ moniae (OD600 0.05) for 90 min, and membrane permeability was bacterial killing was altered in the presence of surfactant phospho- assessed by alkaline phosphatase detection. Helix 1 (residues lipids. Liposomes, composed of DPPC/PG (9:1, w/w) at lipid:pep- 7–22) caused signiﬁcant membrane permeability in a dose-depen- tide ratios of 20:1 or 10:1, were ﬁrst mixed with synthetic peptide dent manner at concentrations as low as 2.5 ␮M (data not shown). followed by incubation with bacteria. Preincubation of liposomes by guest on September 27, 2021

FIGURE 5. Dose-dependent and surfactant inhibition of Klebsiella pneumoniae killing by SP-B. A, In- creasing amounts of synthetic peptide (helix 1 (residues 7–22), N-term helix 1, or helix 1,2) or solvent controls were added to K. pneumoniae (103 CFU) in 100 ␮l of sterile PBS and incubated for6hat37°C. Results (n ϭ 6) are expressed as mean Ϯ p Ͻ 0.05 vs Helix 1; ϩ, p Ͻ ,ء ;SEM 0.05 vs N-term Helix 1. B, SP-B peptides (5.0 ␮M) Helix 1 (residues 7–22) or helix 1,2 were preincubated with DPPC/PG (9:1) liposomes or DPPC liposomes (lipid:protein ratios 20:1 or 10:1) followed by incubation with bacteria. Bacteria were dispersed, plated on blood agar plates, and incubated overnight at 37°C. Re- sults (n ϭ 6) are expressed as mean Ϯ p Ͻ 0.05 vs synthetic peptide ,ء ;SEM lipid-free; ϩ, p Ͻ 0.05 vs DPPC/PG. The Journal of Immunology 421

(DPPC/PG) with the SP-B peptides at 20:1 or 10:1 markedly impaired the ability of helix 1, helix 1,2 (Fig. 5B), and N-term helix 1 (data not shown) to kill bacteria, similar to results obtained for native SP-B (Fig. 1B). Removal of PG from the liposomes signiﬁcantly improved SP-B-mediated bacterial killing (Fig. 5B). PG-mediated inhibition was observed even when the peptides were incubated with bacteria before the addition of liposomes, although the extent of inhibition was much reduced (data not shown). Antimicrobial activity of SP-B peptides against S. aureus To determine whether the SP-B peptides could also kill Gram- positive bacteria, dose-response curves were generated by incubating helix 1 (residues 7–22), N-term helix 1, or helix 1,2 peptide with S. aureus (Fig. 6A). Helix 1,2 killed bacteria at a concentration as low as 0.15 ␮M (55% killing) with maximal bacterial killing at a concentration of 0.6 ␮M, indicating that this peptide was signiﬁcantly more effective than helix 1 (residues 7–22) or N-term helix 1. Helix 2 alone was much less effective at killing bacteria

(Ͻ10% at 2.5 ␮M, data not shown) providing further evidence that Downloaded from helix 1 was required for bacterial killing. Domain-mapping experiments demonstrated that the shorter helix 1 peptide (residues 10– 22) was much less effective at killing S. aureus than the longer peptide (residues 7–22) (Ͻ45% killing at 2.5 ␮M for residues 10–22 compared with Ͼ95% killing for residues 7–22) (Fig. 6B).

These results suggest that the hydrophobic residues tyrosine 7, http://www.jimmunol.org/ cysteine 8, and tryptophan 9 may be important for disrupting membranes of Gram-positive bacteria. SP-B-mediated killing of S. aureus was also inhibited by PG-containing liposomes at a 20:1 lipid:peptide ratio (Fig. 6C); however, decreasing the lipid: peptide ratio to 10:1 dramatically improved bacterial killing (Ͼ90%) with helix 1 (residues 7–22). Both helix 1,2 and helix 1 (residues 7–22) were effective at killing S. aureus (Ͼ95%) in the presence of DPPC vesicles at both the 20:1 and 10:1 lipid: peptide ratios. by guest on September 27, 2021 Identification of amino acids important for bacterial killing The domain-mapping experiments demonstrated that the SP-B peptides containing helix 1 were antimicrobial. To further examine the structural basis for this property and to identify specific residues involved in bacterial killing, amino acid substitutions were introduced into helix 1 in the context of N-term helix 1,2 (i.e., residues 1–37) (Fig. 7). N-term helix 1,2 was previously shown to be the smallest SP-B peptide that promoted surface tension reduc- tion (16). Positively charged amino acids have been shown to be important for the bacteriolytic activity of several antimicrobial FIGURE 6. Dose-dependent and surfactant inhibition of S. aureus peptides (21). To determine whether these residues were also im- killing by SP-B. A, Increasing concentrations of the SP-B peptides (he- portant for the antimicrobial activity of SP-B, positively charged lix 1, N-term helix 1, or helix 1,2) were added to S. aureus (103 CFU) residues located in helix 1 and 2 were systematically substituted in 100 ␮l of sterile PBS and incubated for6hat37°C. Results (n ϭ 6) ,p Ͻ 0.001 vs methanol control; ϩ ,ء ;with uncharged amino acids. We have previously shown that sin- are expressed as mean Ϯ SEM gle alanine or multiple serine substitutions did not alter the sec- p Ͻ 0.001 vs Helix 1,2. B, Dose-dependent killing curves were com- ondary structure of the peptides (16). Substitution of a single pos- pared for Helix 1 peptides, residues 7–22 and 10–22. Results (n ϭ 3) p Ͻ 0.05 vs methanol control. C, The ,ء ;itively charged amino acid (R12, K16, or K24) with alanine had no are expressed as mean Ϯ SEM effect on the antimicrobial activity of SP-B; however, substitution SP-B peptides helix 1 or helix 1,2 (5.0 ␮M) were preincubated with of two or three positively charged residues significantly inhibited DPPC/PG (9:1) liposomes or DPPC liposomes (lipid:protein ratios 20:1 or 10:1) followed by incubation with bacteria. Bacteria were dispersed, killing of K. pneumoniae (Ͻ25%) (Fig. 7) and S. aureus (data not plated on blood agar plates, and incubated overnight at 37°C. Results p Ͻ 0.05 vs lipid-free ,ء ;shown). In particular, substitution of serine for R12 and K16 in (n ϭ 6) are expressed as mean Ϯ SEM helix 1 virtually ablated bacterial killing. These results indicate peptide; ϩ, p Ͻ 0.001 vs DPPC/PG. that at least two positively charged residues in helix 1 are required for the antimicrobial activity of SP-B peptides. the vital dyes Syto 9 and propidium iodide and analyzed by fluo- Effect of SP-B synthetic peptides on bacterial aggregation rescence microscopy (Fig. 8). Helix 1 (residues 7–22) did not in- To assess the ability of synthetic peptides to aggregate bacteria, duce bacterial aggregation but caused a significant increase in pro- helix 1 (residues 7–22) or helix 1,2 were added to K. pneumoniae pidium iodide staining (Fig. 8, D–F) compared with control (A–C).

(OD600 0.1) and incubated for 90 min. Bacteria were stained with Helix 1,2 (residues 7–37) induced bacterial aggregation, but the 422 ANTIMICROBIAL ACTIVITY OF SP-B

FIGURE 7. Effect of cationic amino acid substitutions on SP-B peptide-mediated killing of K. pneumoniae. Individual synthetic SP-B peptides (2.5 ␮M) containing positively charged amino acid substitutions were added to K. pneumoniae (103 CFU) in 100 ␮lof sterile PBS and incubated for6hat37°C. Bacteria were dispersed, plated on blood agar plates, and incubated overnight at 37°C. Results (n ϭ 6) are expressed as .p Ͻ 0.05 vs methanol control ,ء ;mean Ϯ SEM

majority of aggregates were significantly smaller than those in- Hemolytic activity of the SP-B synthetic peptides Downloaded from duced by native SP-B (mean aggregate area, 50 Ϯ 10 ␮m2; p Ͻ 0.001) (Fig. 8G-I). In a few fields, larger bacterial aggregates were To determine the specificity of SP-B peptides for prokaryotic cell observed with sizes similar to those induced by native SP-B (Fig. membranes, N-term helix 1,2, helix 1 (residues 7–22), helix 1 (res- 8, J–L). Virtually all of the bacteria within the aggregates were idues 10–22; data not shown), helix 1,2, N-term helix 1, and melit- positive for propidium iodide staining consistent with dead/dying tin were incubated with hRBC for 1 h. All of the SP-B peptides bacteria (Fig. 8K). Helix 1,2 exhibited similar activity toward other tested exhibited very low levels of hemolytic activity compared http://www.jimmunol.org/ strains of bacteria including P. aeruginosa and group B strepto- with melittin (Ͻ15% hemolysis at the highest concentration) (Fig. coccus (data not shown). 9). Incubation of A549 cells with 5 ␮M helix 1 (residues 7–22) for by guest on September 27, 2021

FIGURE 8. Effect of synthetic SP-B peptides on bacterial aggregation and membrane permeabilization.

K. pneumoniae (OD600 0.1) were incubated with the SP-B peptides (10 ␮M) helix 1 (residues 7–22) (D–F) or helix 1,2 (G–L) or solvent (methanol) controls (A–C) for 90 min at 37°C. Bacteria were stained with the vital dyes Syto 9 and propidium iodide and analyzed by ﬂu- orescence microscopy to assess bacterial aggregation and viability. Results are representative of four separate experiments. Helix 1,2 (residues 7–37) induced very small aggregates (G–I); however, in a few ﬁelds, larger bacterial aggregates were detected (J–L). Bar, 20 ␮m. The Journal of Immunology 423

FIGURE 9. Hemolytic activity of SP-B peptides. Synthetic peptides N-term helix 1,2, helix 1,2, N-term helix 1, and helix 1 (residues 7–22) were incubated with 4% hRBC for1hat37°C. Methanol-treated controls were included in all experiments and had no effect on hemolysis. Results are the mean of four separate exper- .p Ͻ 0.001 vs melittin ,ء ;iments Ϯ SEM

1 h resulted in death of 33.3 Ϯ 6.2% of cells; incubation of cells Discussion with the solvent (methanol) control resulted in death of 28 Ϯ 5.8% SP-B is a member of the SAPLIPs, which include the potent an- of cells ( p ϭ 0.3977). timicrobial peptides NK-lysin, granulysin, and amoebapore. SAP- Downloaded from LIPs are characterized by a conserved disulfide bond pattern and Effect of surfactant phospholipids on SP-B peptide-mediated likely share a similar tertiary structure. All SAPLIPs interact with bacterial killing lipids and several, including the antimicrobial peptides, have mem- Helix 1 (residues 7–22) was more effective at killing S. aureus than branolytic activity. The interaction of peptides with membranes is K. pneumoniae in the presence of lipids compared with helix 1,2 or mediated in part by cationic residues usually located in the polar N-term helix 1 (Figs. 5B and 6C); in particular, helix 1 killed S. face of an amphipathic helix. SP-B interacts with the surface of the http://www.jimmunol.org/ aureus much more effectively than helix 1,2 in the presence of lipid bilayer via four or five amphipathic ␣ helices (20). Positively Ͼ DPPC/PG ( 90% killing for helix 1 at the 10:1 ratio compared charged amino acids, located predominantly in the N-terminal do- Ͻ with 5% killing for helix 1,2) (Fig. 6C). We next determined main of SP-B, facilitate interaction of the mature peptide with the whether helix 1 could augment the ability of native surfactant to negatively charged head groups of PG (22–24). Domain-mapping kill bacteria. Bronchoalveolar lavage was performed on wild-type experiments demonstrated that a cationic peptide corresponding to mice, cells were removed using low-speed centrifugation, and sur- helix 1 (residues 7–22) was sufficient to lyse negatively charged factant phospholipids and associated proteins were pelleted at vesicles (16). The results of the current study indicate that native 18,000 ϫ g for 15 min. Increasing amounts of helix 1 peptide were

SP-B and synthetic peptide derivatives containing helix 1 killed by guest on September 27, 2021 ␮ added to 0.75 g of total surfactant lipid followed by incubation Gram-positive and Gram-negative bacteria in vitro. with K. pneumoniae or S. aureus (103 CFU) for6hat37°C. In the present study, native SP-B, isolated from human BALF, Bacteria were plated on blood agar plates, and colonies were killed clinical isolates of Gram-positive and Gram-negative bac- counted after 18 h. In the presence of native surfactant, SP-B helix teria in a dose-dependent manner. Antimicrobial activity was de- 1 killed both K. pneumoniae and S. aureus at concentrations of tected at concentrations between 0.06 and 1.0 ␮M, comparable to 5–10 ␮M (Fig. 10). other potent antimicrobial SAPLIP peptides and other well-characterized ␣-helical, cationic peptides (12, 25–27). Native SP-B killed bacteria by permeabilizing the bacterial cell membrane, as indicated by detection of alkaline phosphatase activity and increased propidium iodide staining. Peptide-mapping experiments demonstrated that the antimicrobial activity of SP-B mapped with the lytic activity to helix 1 (residues 7–22). This finding supports the observation of Kaser and Skouteris (9), who noted that residues 12–34 of SP-B are 68% homologous to residues 48–72 of the frog peptide antibiotic dermaseptin bI. In addition to direct bacterial killing, SP-B also induced significant aggregation of Gram-positive and Gram-negative bacteria. Bacterial aggregation facilitated by the collectins SP-A and SP-D likely plays a role in enhancement of phagocytosis, complement activation, and/or inhibition of microbial colonization and invasion (3). SP-A and SP-D bind polysaccharides located on the surface of bacteria through their C-terminal carbohydrate recognition domains. Domain-mapping experiments of SP-B indicated that, al- FIGURE 10. SP-B helix 1 kills bacteria in the presence of native sur- though helix 1 was sufficient for bacterial killing, aggregation re- factant. Surfactant was isolated by high-speed centrifugation of cell-free quired both helix 1 and helix 2. This finding agrees well with a BALF obtained from FVBN mice (6–8 wk old). Increasing concentrations previous study (16) that implicated helix 2 in membrane cross- of SP-B helix 1 (residues 7–22) were added to 0.75 ␮g of total surfactant phospholipid and incubated with bacteria (103 CFU of K. pneumoniae or S. linking (aggregation) and fusion (promoting lipid transfer between aureus). Samples were incubated for6hat37°C, dispersed, plated on lipid bilayers and the surface active monolayer): SP-B peptides blood agar plates, and incubated overnight to obtain colony counts. Results anchored to separate membranes by helix 1 may cross-link mem- ;(p Ͻ 0.05 vs methanol control. branes by interacting through helix 2 (peptide-peptide interaction ,ء ;n ϭ 3) are expressed as mean Ϯ SEM) 424 ANTIMICROBIAL ACTIVITY OF SP-B alternatively, the SP-B peptide may form a “bridge” in which helix and 4) spontaneous deinsertion of peptide with redistribution on 1 interacts with one membrane and helix 2 interacts with a separate both sides of the membrane, permitting access of peptides to in- membrane (peptide-lipid interaction). It is interesting to note that tracellular targets. Accumulation of cationic SP-B peptides at lev- only a fraction of the bacteria aggregated by native SP-B were els sufficient to initiate translocation and membrane permeabiliza- killed (Fig. 2), whereas virtually all of the bacteria aggregated by tion would be critically dependent on electrostatic interactions and peptide helix 1,2 were stained by propidium iodide (Fig. 8). This interfacial hydrophobicity. Only negatively charged membranes suggests that the synthetic peptide aggregates bacteria through “le- would attract enough of the smaller, cationic, synthetic peptides to thal” domains, presumably helix 1, whereas native SP-B may in- form permeabilizing oligomers; in contrast, native SP-B, which is duce bacterial aggregation via multiple motifs, some of which lack intrinsically oligomerized, may be competent to permeabilize both killing activity. However, although SP-B clearly promoted bacte- anionic and zwitterionic membranes, even at low protein densities. rial aggregation in vitro, the importance of this property for bac- The shortest membranolytic SP-B peptide, helix 1, was more terial killing and/or clearance in vivo is less certain. resistant to inhibition by phospholipids than the native peptide and SP-B is very hydrophobic and is likely always associated with retained significant antimicrobial activity in the presence of native surfactant phospholipids in vivo. Although lipid-free SP-B was surfactant. A synthetic peptide containing helix 2 (helix 1,2) killed bactericidal in vitro, this activity was dramatically inhibited in the bacteria at lower concentrations than helix 1 alone and aggregated presence of surfactant phospholipids, particularly PG. We previ- bacteria similarly to the native peptide; however, this peptide was ously reported (28) that the content of SP-B in the alveolar spaces more sensitive to lipid inhibition. The lipid vesicle aggregates in- was 5–6 ␮g, and the total surfactant phospholipid content was 275 duced by native SP-B or helix 1,2 may hide a significant fraction

␮g (estimated for a 25-g mouse at 6–8 wk of age). Thus the of the peptide, thereby decreasing transfer to the bacterial mem- Downloaded from lipid:protein ratios used in the current study (20:1) were much brane. Because helix 1 does not aggregate membranes, it may be lower than the estimated ratio in vivo (50:1). These data strongly fully exposed on the surface of the vesicles where it can be readily suggest that PG will inhibit the antimicrobial activity of SP-B in transferred to bacterial membranes. vivo. Furthermore, lipid-free SP-B exhibited hemolytic activity Residues 7–9 of SP-B were required for efﬁcient killing of comparable to melittin, and this activity was completely inhibited Gram-positive bacteria but not Gram-negative bacteria. This dif-

in the presence of surfactant phospholipids, indicating the impor- ference may be related to the intrinsically different structure of the http://www.jimmunol.org/ tance of maintaining native SP-B in a lipid-associated form. We target membranes of these microorganisms. Permeabilization of cannot exclude the possibility that, in vivo, some SP-B may exist Gram-negative bacteria would require translocation through the in microdomains that are enriched in DPPC and are relatively poor external LPS containing envelope and the periplasmic space before in PG content. Such a microenvironment would minimize the he- reaching the target plasma membrane. Both the external envelope molytic activity of SP-B while preserving at least some of its an- and the plasma membrane have anionic surfaces and could accu- tibiotic properties. We also cannot exclude the possibility that na- mulate peptide through electrostatic affinity. The presence of com- tive SP-B may be proteolytically cleaved into smaller peptide peting anionic membranes (DPPC/PG vesicles) would inhibit par- fragments, similar to the synthetic peptide derivatives described in titioning of peptides into both layers. Gram-positive bacteria such the current study, i.e., peptides that retain antimicrobial activity in as S. aureus have a single membrane with a thick external, nega- by guest on September 27, 2021 the presence of surfactant and have little or no hemolytic activity. tively charged wall containing peptidoglycan and teichoic acid. The generation of antimicrobial peptides from precursor proteins Electrostatic interactions would facilitate peptide accumulation at has been reported previously. For example, buforin I, a peptide that the membrane surface, but penetration of the phospholipid bilayer is important for innate host defense of the intestinal epithelium, is would be dependent on interfacial hydrophobicity, conferred pre- generated by enzymatic cleavage of the non-antimicrobial precur- dominantly by the aromatic side chains of Tyr7 and Trp9. This sor protein histone H2A (29). SP-B peptide fragments could be model would explain why 1) lipid vesicles are less able to inhibit generated by a similar process; alternatively, SP-B peptide frag- the antibiotic activities of SP-B peptides toward S. aureus than K. ments could be generated within alveolar macrophages following pneumoniae, and 2) removal of aromatic residues Tyr7 and Trp9 uptake from the airspaces. However, the results of studies in trans- produced a substantial decrease in the anti-staphylococcal proper- genic mice are not consistent with the generation of bactericidal ties of helix 1. Consistent with this model, Serrano et al. (31) peptides or specialized lipid microdomains. Increased expression recently demonstrated that residues 7–9 exhibited the highest af- of SP-B in transgenic mice did not enhance bacterial killing of P. finity for phospholipid interfaces of any motif in SP-B. aeruginosa, group B streptococcus (10), or K. pneumoniae (cur- In summary, although a significant role for endogenous SP-B in rent study), and, importantly, susceptibility to bacterial infection innate host defense of the lung may be limited, synthetic peptides was not increased in mice in which the concentrations of SP-B in derived from native SP-B may be very useful as antimicrobial the airspaces was decreased by 50% (10). Thus, although a role for agents. SP-B peptides encompassing helix 1 (helix 1, N-term helix native SP-B in host defense remains a formal possibility, the ex- 1, and helix 1,2) exhibited potent antimicrobial activity against perimental evidence to support this hypothesis is currently lacking. clinical isolates of K. pneumoniae, S. aureus, group B streptococ- Synthetic peptide derivatives of SP-B exhibited little to no he- cus, and P. aeruginosa at low peptide concentrations in vitro. The molytic activity and selectively lysed bacterial membranes. The properties of bacterial killing in the presence of surfactant phos- difference in membrane selectivity between native SP-B and the pholipids and selectivity for bacterial membranes suggest that he- peptide derivatives could be due to the mechanism of action that lix 1 (residues 7–22) may be useful as an adjunct for treatment of has been proposed for a large number of cationic antibiotic pep- bacterial pneumonias. tides. According to the carpet model proposed by Shai (30), cat- ␣ ionic, amphipathic, -helical peptides act on bacterial membranes Acknowledgments in four main steps including 1) interfacial partitioning with accu- We thank Chenxia Duan and Richard Papes for technical assistance. mulation of monomeric peptides on the target membrane (limiting step); 2) peptide rearrangement, usually via oligomerization; 3) membrane permeabilization/depolarization associated with adop- Disclosures tion of a transient transmembrane orientation of peptide oligomers; The authors have no financial conflict of interest. The Journal of Immunology 425

References 16. Ryan, M. A., X. Y. Qi, A. G. Serrano, M. Ikegami, J. Perez-Gil, J. Johansson, and T. E. Weaver. 2005. Mapping and analysis of the lytic and fusogenic domains of 1. Clark, J. C., S. E. Wert, C. J. Bachurski, M. T. Stahlman, B. R. Stripp, surfactant protein B. Biochemistry 44: 861–872. T. E. Weaver, and J. A. Whitsett. 1995. Targeted disruption of the surfactant 17. Shen, H. Q., C. X. Duan, Z. Y. Li, and Y. Suzuki. 1997. Effects of proteinosis protein B gene disrupts surfactant homeostasis, causing respiratory failure in surfactant proteins on the viability of rat alveolar macrophages. Am. J. Respir. newborn mice. Proc. Natl. Acad. Sci. USA 92: 7794–7798. Crit. Care Med. 156: 1679–1687. 2. Melton, K. R., L. L. Nesslein, M. Ikegami, J. W. Tichelaar, J. C. Clark, J. A. Whitsett, and T. E. Weaver. 2003. SP-B deficiency causes respiratory failure 18. Horowitz, A. D., B. Elledge, J. A. Whitsett, and J. E. Baatz. 1992. Effects of lung in adult mice. Am. J. Physiol. 285: L543–L549. surfactant proteolipid SP-C on the organization of model membrane lipids—a 3. Crouch, E. C. 1998. Collectins and pulmonary host defense. Am. J. Respir. Cell fluorescence study. Biochim. Biophys. Acta 1107: 44–54. Mol. Biol. 19: 177–201. 19. Bartlett, G. R. 1959. Phosphorous assay in column chromatography. J. Biol. 4. Ikegami, M., C. L. Na, T. R. Korfhagen, and J. A. Whitsett. 2005. Surfactant Chem. 234: 466–468. protein D influences surfactant ultrastructure and uptake by alveolar type II cells. 20. Andersson, M., T. Curstedt, H. Jornvall, and J. Johansson. 1995. An amphipathic Am. J. Physiol. 288: L552–L561. helical motif common to tumourolytic polypeptide NK-lysin and pulmonary sur- 5. Kabha, K., J. Schmegner, Y. Keisari, H. Parolis, J. Schlepper-Schaefer, and factant polypeptide SP-B. FEBS Lett. 362: 328–332. I. Ofek. 1997. SP-A enhances phagocytosis of Klebsiella by interaction with 21. Shai, Y. 1999. Mechanism of the binding, insertion and destabilization of phos- capsular polysaccharides and alveolar macrophages. Am. J. Physiol. 16: pholipid bilayer membranes by ␣-helical antimicrobial and cell non-selective L344–L352. membrane-lytic peptides. Biochim. Biophys. Acta 1462: 55–70. 6. Mcneely, T. B., and J. D. Coonrod. 1993. Comparison of the opsonic activity of 22. Baatz, J. E., B. Elledge, and J. A. Whitsett. 1990. Surfactant protein SP-B induces human surfactant protein A for staphylococcus-aureus and streptococcus-pneu- ordering at the surface of model membrane bilayers. Biochemistry 29: moniae with rabbit and human macrophages. J. Infect. Dis. 167: 91–97. 6714–6720. 7. Ofek, I., A. Mesika, M. Kalina, Y. Keisari, R. Podschun, H. Sahly, D. Chang, D. McGregor, and E. Crouch. 2001. Surfactant protein D enhances phagocytosis 23. Perez-Gil, J., C. Casals, and D. Marsh. 1995. Interactions of hydrophobic lung and killing of unencapsulated phase variants of Klebsiella pneumoniae. Infect. surfactant proteins SP-B and SP-C with dipalmitoylphosphatidylcholine and di- Immun. 69: 24–33. palmitoylphosphatidylglycerol bilayers studied by electron spin resonance spectroscopy. Biochemistry 34: 3964–3971. 8. Wu, H. X., A. Kuzmenko, S. J. Wan, L. Schaffer, A. Weiss, J. H. Fisher, Downloaded from K. S. Kim, and F. X. McCormack. 2003. Surfactant proteins A and D inhibit the 24. Vandenbussche, G., A. Clercx, M. Clercx, T. Curstedt, J. Johansson, H. Jornvall, growth of Gram-negative bacteria by increasing membrane permeability. J. Clin. and J. M. Ruysschaert. 1992. Secondary structure and orientation of the surfactant Invest. 111: 1589–1602. protein SP-B in a lipid environment—a Fourier transform infrared spectroscopy 9. Kaser, M. R., and G. G. Skouteris. 1997. Inhibition of bacterial growth by syn- study. Biochemistry 31: 9169–9176. thetic SP-B1-78 peptides. Peptides 18: 1441–1444. 25. Andra, J., O. Berninghausen, J. Wulfken, and M. Leippe. 1996. Shortened amoe- 10. Akinbi, H. T., H. Bhatt, W. M. Hull, and T. E. Weaver. 1999. Altered surfactant bapore analogs with enhanced antibacterial and cytolytic activity. FEBS Lett. protein B levels in transgenic mice do not affect clearance of bacteria from the 385: 96–100. lungs. Pediatr. Res. 46: 530–534. 26. Benincasa, M., B. Skerlavaj, R. Gennaro, A. Pellegrini, and M. Zanetti. 2003. In http://www.jimmunol.org/ 11. Patthy L. 1991. Homology of the precursor of pulmonary surfactant-associated vitro and in vivo antimicrobial activity of two ␣-helical cathelicidin peptides and protein SP-B with prosaposin and sulfated glycoprotein-1. J. Biol. Chem. 266: of their synthetic analogs. Peptides 24: 1723–1731. 6035–6037. 27. Matsuzaki, K., K. Sugishita, M. Harada, N. Fujii, and K. Miyajima. 1997. Inter- 12. Bruhn, H., B. Riekens, O. Berninghausen, and M. Leippe. 2003. Amoebapores actions of an antimicrobial peptide, magainin 2, with outer and inner membranes and NK-lysin, members of a class of structurally distinct antimicrobial and cy- of Gram-negative bacteria. Biochim. Biophys. Acta 1327: 119–130. tolytic peptides from protozoa and mammals: a comparative functional analysis. 28. Nesslein, L. L., K. R. Melton, M. Ikegami, C. L. Na, S. E. Wert, W. R. Rice, Biochem. J. 375: 737–744. J. A. Whitsett, and T. E. Weaver. 2005. Partial SP-B deficiency perturbs lung 13. Ernst, W. A., S. Thoma-Uszynski, P. Teitelbaum, C. Ko, D. A. Hanson, function and causes air space abnormalities. Am. J. Physiol. 288: L1154–L1161. C. Clayberger, A. M. Krensky, M. Leippe, B. R. Bloom, T. Ganz, and R. L. Modlin. 2000. Granulysin, a T cell product, kills bacteria by altering mem- 29. Kim, H. S., H. Yoon, I. Minn, C. B. Park, W. T. Lee, M. Zasloff, and S. C. Kim. brane permeability. J. Immunol. 165: 7102–7108. 2000. Pepsin-mediated processing of the cytoplasmic histone H2A to strong an- 14. Hecht, O., N. A. Van Nuland, K. Schleinkofer, A. J. Dingley, H. Bruhn, timicrobial peptide buforin I. J. Immunol. 165: 3268–3274. by guest on September 27, 2021 M. Leippe, and J. Grotzinger. 2004. Solution structure of the pore-forming pro- 30. Shai, Y. 2002. Mode of action of membrane active antimicrobial peptides. tein of Entamoeba histolytica. J. Biol. Chem. 279: 17834–17841. Biopolymers 66: 236–248. 15. Poulain, F. R., L. Allen, M. C. Williams, R. L. Hamilton, and S. Hawgood. 1992. 31. Serrano, A. G., M. Ryan, T. E. Weaver, and J. Perez-Gil. Critical structure- Effects of surfactant apolipoproteins on liposome structure—implications for tu- function determinants within the N-terminal region of pulmonary surfactant pro- bular myelin formation. Am. J. Physiol. 262: L730–L739. tein SP-B. Biophys. J. In press.